[gmx-users] problem with g_cluster

sridhar at cdfd.org.in sridhar at cdfd.org.in
Thu Jun 23 03:54:34 CEST 2005


Hi GMX users,

I have a problem with g_cluster. I am using g_cluster to cluster my
trajectory of 20000 structures
(20000ps). I am using the matrix file as input using option -dm. The
program gave error indicating memory fault initially. Later I ran the same
program using -skip 1000 option, thereby effectively reading only 20
structures, to check if it runs. Surprisingly the same problem repeats
even with this option. I am running on a intel(XEON dual
processor) machine with 2GB RAM and I think
it is sufficient for the process. I cannot figure out the problem.
Waiting for your suggessions.
I am attaching the output messages of the program below
#############################################################################
                          :-)   G  R  O  M  A  C  S  (-:

              GROningen Mixture of Alchemy and Childrens' Stories

                             :-)   VERSION 3.2.1  (-:


      Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
       Copyright (c) 1991-2000, University of Groningen, The Netherlands.
             Copyright (c) 2001-2004, The GROMACS development team,
            check out http://www.gromacs.org for more information.

         This program is free software; you can redistribute it and/or
          modify it under the terms of the GNU General Public License
         as published by the Free Software Foundation; either version 2
             of the License, or (at your option) any later version.

                               :-)   g_cluster  (-:

Option     Filename  Type         Description
------------------------------------------------------------
  -f       traj.xtc  Input, Opt.  Generic trajectory: xtc trr trj gro g96
pdb
  -s /home/sridhar/CYP_NEWDYN/b4md_1cyp_WT.tpr  Input, Opt!
                                   Structure+mass(db): tpr tpb tpa gro g96
                                   pdb xml
  -n /home/sridhar/CYP_NEWDYN/b4ion_1cyp_WT.ndx  Input, Opt!  Index file
 -dm ./RMSDMAT_WT.xpm  Input, Opt!  X PixMap compatible matrix file
  -o rmsd-clust.xpm  Output       X PixMap compatible matrix file
  -g cluster_WT.log  Output       Log file
-dist rmsd-dist.xvg  Output, Opt. xvgr/xmgr file
 -ev   rmsd-eig.xvg  Output, Opt. xvgr/xmgr file
 -sz clust_size_WT.xvg  Output, Opt! xvgr/xmgr file
 -tr clust_trans_WT.xpm  Output, Opt! X PixMap compatible matrix file
-ntr clust_Ntrans_WT.xvg  Output, Opt! xvgr/xmgr file
-clid clust_it_WT.xvg  Output, Opt! xvgr/xmgr file
 -cl clusters_WT.pdb  Output, Opt! Generic trajectory: xtc trr trj gro g96
pdb

      Option   Type  Value  Description
------------------------------------------------------
      -[no]h   bool     no  Print help info and quit
       -nice    int      0  Set the nicelevel
          -b   time     -1  First frame (ps) to read from trajectory
          -e   time     -1  Last frame (ps) to read from trajectory
         -dt   time     -1  Only use frame when t MOD dt = first time (ps)
         -tu   enum     ps  Time unit: ps, fs, ns, us, ms, s, m or h
      -[no]w   bool     no  View output xvg, xpm, eps and pdb files
  -[no]dista   bool     no  Use RMSD of distances instead of RMS deviation
    -nlevels    int     40  Discretize RMSD matrix in # levels
   -keepfree    int     -4  if >0 # levels not to use when coloring
clusters;
                            if <0 nlevels/-keepfree+1 levels will not be
used
     -cutoff   real    0.1  RMSD cut-off (nm) for two structures to be
                            neighbor
    -[no]fit   bool    yes  Use least squares fitting before RMSD
calculation
        -max   real     -1  Maximum level in RMSD matrix
       -skip    int   1000  Only analyze every nr-th frame
     -[no]av   bool    yes  Write average iso middle structure for each
                            cluster
        -wcl    int      0  Write all structures for first # clusters to
                            numbered files
        -nst    int      1  Only write all structures if more than # per
                            cluster
     -rmsmin   real      0  minimum rms difference with rest of cluster
for
                            writing structures
     -method   enum jarvis-patrick  Method for cluster determination:
                            linkage, jarvis-patrick, monte-carlo,
                            diagonalization or gromos
  -minstruct    int      1  Minimum number of structures in cluster for
                            coloring in the xpm file
 -[no]binary   bool     no  Treat the RMSD matrix as consisting of 0 and
1,
                            where the cut-off is given by -cutoff
          -M    int     10  Number of nearest neighbors considered for
                            Jarvis-Patrick algorithm, 0 is use cutoff
          -P    int      3  Number of identical nearest neighbors required
to
                            form a cluster
       -seed    int   1993  Random number seed for Monte Carlo clustering
                            algorithm
      -niter    int  10000  Number of iterations for MC
         -kT   real  0.001  Boltzmann weighting factor for Monte Carlo
                            optimization (zero turns off uphill steps)


Warning: cannot write cluster structures without reading trajectory
         ignoring option -cl clusters_WT.pdb

Back Off! I just backed up cluster_WT.log to ./#cluster_WT.log.5#
Using jarvis-patrick method for clustering
Will use P=3, M=10 for determining the neighbors

Reading rms distance matrix 100%
Converted a 20001x20001 matrix with 80 levels to reals
The RMSD ranges from 0.0514 to 0.451 nm
Average RMSD is 0.0419409
Number of structures for matrix 20001
Energy of the matrix is 1854.48 nm
WARNING: rmsd minimum 0 is below lowest rmsd value 0.0514
Linking structures cluster.sh: line 2:  5007 Killed
g_cluster -dm ./RMSDMAT_$abc.xpm -s
/home/sridhar/CYP_NEWDYN/b4md_1cyp_$abc.tpr -n
/home/sridhar/CYP_NEWDYN/b4ion_1cyp_$abc.ndx -g cluster_$abc.log -sz
clust_size_$abc.xvg -tr clust_trans_$abc.xpm -ntr clust_Ntrans_$abc.xvg
-clid clust_it_$abc.xvg -cl clusters_$abc.pdb -av -method jarvis-patrick
-nice 0 -skip 1000
##########################################################################


thanks and waiting for your suggessions.

M. Sridhar Acharya
Senior Research Fellow
Lab. of Computational Biology
Centre for DNA Fingerprinting and Diagnostics
Hyderabad. 500076
Ph: +91-40-27151344-1207
Fax: +91-40-27155479




More information about the gromacs.org_gmx-users mailing list